AI-Powered Dust Storm Movement Prediction System Using Satellite Imagery

The prediction of the path of the dust storms is not an easy task and the same has been achieved in this article. For the task to be completed FCN, fully convolutional network-based AI model has been developed and trained appropriately. The model is well built in such a way that the accuracy is very appreciable. The model has presented us with almost 100 percent accuracy with training and 93.2% accuracy with the validation. The model can be integrated into the early warning system which will provide the alerts in time. With that kind of alerts, one could take informed decisions and mitigate all the side effects and impacts. This solution can help in preventing damages to the people, infrastructure, and even agriculture lands which is heart of any nation. The work has been completed with Intel oneAPI as the platform, and the solution is found be stable in all working conditions.